BACKGROUND: Procalcitonin (PCT) levels can be used to predict bacteremia and DNAemia in patients with sepsis. In this study, the diagnostic accuracy of PCT in predicting blood culture (BC) results and DNAemia, as detected by real-time PCR (RT-PCR), was compared with that of other markers of inflammation commonly evaluated in patients with suspected sepsis, such as C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), and white blood cell (WBC) count. METHODS: A total of 571 patients for whom BC, blood RT-PCR, PCT, CRP, ESR, and WBC count were requested for laboratory diagnosis of sepsis were included in the study. Receiver operating characteristic curve analysis was performed to compare the ability of the above biomarkers to predict BC and blood RT-PCR results. RESULTS: A total of 108 pathogens were identified by BC (79 pathogens, 14.5% positive rate) and/or RT-PCR (90 pathogens, 16.5% positive rate), after exclusion of 26 contaminated samples. The PCT areas under the curve (AUCs) in predicting BC (0.843; 95% CI 0.796-0.890; p < 0.0001) and RT-PCR (0.916; 95% CI 0.888-0.945; p < 0.0001) results were significantly greater than AUCs found for CRP, ESR, and WBC count. CONCLUSIONS: PCT showed a better diagnostic accuracy than CRP, ESR, and WBC count in predicting DNAemia and bacteremia in patients with suspected sepsis.
BACKGROUND: Procalcitonin (PCT) levels can be used to predict bacteremia and DNAemia in patients with sepsis. In this study, the diagnostic accuracy of PCT in predicting blood culture (BC) results and DNAemia, as detected by real-time PCR (RT-PCR), was compared with that of other markers of inflammation commonly evaluated in patients with suspected sepsis, such as C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), and white blood cell (WBC) count. METHODS: A total of 571 patients for whom BC, blood RT-PCR, PCT, CRP, ESR, and WBC count were requested for laboratory diagnosis of sepsis were included in the study. Receiver operating characteristic curve analysis was performed to compare the ability of the above biomarkers to predict BC and blood RT-PCR results. RESULTS: A total of 108 pathogens were identified by BC (79 pathogens, 14.5% positive rate) and/or RT-PCR (90 pathogens, 16.5% positive rate), after exclusion of 26 contaminated samples. The PCT areas under the curve (AUCs) in predicting BC (0.843; 95% CI 0.796-0.890; p < 0.0001) and RT-PCR (0.916; 95% CI 0.888-0.945; p < 0.0001) results were significantly greater than AUCs found for CRP, ESR, and WBC count. CONCLUSIONS: PCT showed a better diagnostic accuracy than CRP, ESR, and WBC count in predicting DNAemia and bacteremia in patients with suspected sepsis.